[Poohbah]

The EMP effects for a small nuclear explosion at ground level extend only about as far as the blast effects. This would be just as true in a city. The forest of metal in the city will serve to attenuate the EMP and reflect it back into ground zero just the same as the city environment tends to attenuate TV signals.

TV signals are many orders of magnitude lower in power than the emissions from even a small nuclear fireball. ANYTHING is transparent to electromagnetic radiation, given sufficient signal amplitude...

Basically, if the fireball touches it, it's going to reradiate. If it's touching something the fireball touches, it will reradiate.

[Dan Evans]

EMP is not significant for small bombs. If the EMP is big enough to fry your transistors, the blast will get you anyway:

... The charge separation persists for only a few tens of microseconds, making the emission power some 100 gigawatts. The field strengths for ground bursts are high only in the immediate vicinity of the explosion. For smaller bombs they aren't very important because they are strong only where the destruction is intense anyway. With increasing yields, they reach farther from the zone of intense destruction. With a 1 Mt bomb, they remain significant out to the 2 psi overpressure zone (5 miles).

http://nuclearweaponarchive.org/Nwfaq/Nfaq5.html#nfaq5.5